It has been shown that tumors have developed numerous ways to escape tumor specific immune responses. These mechanisms will ultimately not only enhance tumor growth, but also impair the effect of immune based therapies in cancer. Myeloid derived suppressor cells represent a recently identified cell population, which has been shown to impair tumor specific immune responses both in mice and human. MDSC can be divided in two subtypes (namely monocytic and granulocytic MDSC. We have been able to examine and define the regulation and function of murine MDSC by IFN-gamma and studied MDSC in different HCC models 1. We described a new IFN-gamma -dependent regulator mechanism of the suppressor function of MDSC. While IFN-gamma blockade impairs the suppressor function of monocytic CD11b+Gr-1dull/int. it has opposing effects on granulocytic CD11b+Gr-1high MDSC. Co-culture of CD11b+Gr-1high granulocytic MDSC with antigen-stimulated T cells and simultaneous blockade of IFN-gamma by the use of anti-IFN-gamma blocking antibody, IFN-gamma-/- effector T cells, IFN-gammaR-/- MDSC or STAT1 -/- MDSC led to up-regulation of Bcl2a1 in CD11b+Gr-1high granulocytic MDSC, improved survival of granulocytic subpopulation during MDSC-T cell interaction and enhanced suppressor function. Molecular studies revealed that GM-CSF released by antigen-stimulated CD8+T cells induced Bcl2a1 up-regulation, which was repressed in the presence of IFN-gamma by a direct interaction of phosphorylated STAT-1 with the Bcl2a1 promotor. Bcl2a1 overexpressing MDSC not only demonstrated prolonged survival in vitro and enhanced suppressor function in vitro but also showed improved suppressor function in vivo. Our data suggest that IFN-gamma/ STAT1 -dependent regulation of Bcl2a1 regulates survival and thereby suppressor function of CD11b+Gr-1high MDSC. 2. Myeloid derived suppressor cells (MDSC) are immature myeloid cells with immunosuppressive activity. They accumulate in tumor-bearing mice and humans with different types of cancer, including hepatocellular carcinoma (HCC). We examined the biology of MDSC in murine HCC models and to identify a model, which mimics the human disease. An accumulation of MDSC was found in mice with HCC irrespectively of the model tested. Transplantable tumors rapidly induced systemic recruitment of MDSC, in contrast to slow-growing DEN-induced or MYC-expressing HCC, where MDSC numbers only increased intra-hepatically in mice with advanced tumors. MDSC derived from mice with subcutaneous tumors were more suppressive than those from mice with DEN-induced HCC. Enhanced expression of genes associated with MDSC generation (GM-CSF, VEGF, IL-6, IL-1beta) and migration (MCP-1, KC, S100A8, S100A9) was observed in mice with subcutaneous tumors. In contrast, only KC levels increased in mice with DEN-induced HCC. Both KC and GM-CSF over-expression or anti-KC and anti-GM-CSF treatment controlled MDSC frequency in mice with HCC. Finally, the frequency of MDSC decreased upon successful anti-tumor treatment with sorafenib. 3. We have stated to study murine models of NASH and HCC. Currently we are evaluating how NASH affects immunosurveillance and potentially promotes tumor development by affecting anti-tumor immunity.